Hanging DB2 threads

[Marley777]

Can anyone help us with ways to deal with hanging DB2 threads. It's as if DataStage is not cleaning up it's resources after an abort. The only way we can kill these threads is to bring db2 down. Is there a better way? Why do threads hang when using DataStage?

UNIX 5.3
DB2 Version 8
DataStage 7.5.1


Thanks for your time!!!

[gbusson]

it may be a DB2 problem, not a Datastage problem.

do the abort occur during the load?

[track_star]

How bout sendint a ctrl-d to the pipe?

[Marley777]

Hello All thanks for your response so far.

right now I'm trying ctrl-shift-v and then a ctrl-d then pipe.


What is sendint?

[track_star]

Typo....sorry. I meant to type sending.

[Marley777]

Hi,

it looks like we are talking about the same thing. Let me try this and see what happens. I will let everyone know.

Thanks

[Marley777]

Hello All,

using this format echo [ctrl-shift-v then crtl-d] > pipename

I then type echo^D>ordb221340180a0324b9_.009


but nothing happens. My command promt never comes back either.
am I using the command correctly?


the file name above is in my scratch datastage directory.

prw-rw-rw- 1 xmm13 staff 0 Oct 20 13:35 ordb221340180a0324b9_.0
09

[vmcburney]

Do you have a job that is trying to insert and update to the same table at the same time via seperate stages? Do you have a job using the Load - Import feature? I have found it extremely easy to lock a DB2 table, easier then any other RDBMS I have used, we often get tables locked up via read only queries from Quest! A failed import can leave the table locked up and a DBA needs to perform the cleanup manually.

Have a look at the options in your DB2 stage to see if you can set some auto cleanup settings. Try to split your inserts and updates into seperate jobs. If a job locks up get everyone to shut down their query sessions such as DB2 Control Center and Quest.